INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 leaked

-0.62

 lash

-0.61

 bidder

-0.60

Bei

-0.59

Bou

-0.59

naissance

-0.58

 booklet

-0.58

 complied

-0.57

 consulted

-0.57

 intact

-0.57

POSITIVE LOGITS

ï¸

0.88

TPPStreamerBot

0.85

oppable

0.75

FTWARE

0.75

prototype

0.73

ociate

0.67

rontal

0.66

uten

0.65

feature

0.64

ignt

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.